Hey everyone, I am a little stuck and really hope I can get some help here
Situation, one of my HDDs died and luckily I have a slight older Backup made by UrBackup. So far so good, it’s quite a lot data and it was on a single 3TB HDD but partioned into 2 Volumes (1TB + 2TB) They were on a Windows machine and got the Symbols B: and D: (Since C: is the System)
That 3TB HDD is no longer available and I don’t plan to replace it, for now I want to get my Files back from the Backup on an external Drive (5TB HDD). I have one non-incremental and some incremental backups for all Drives, but relevant are only 2 (B: & D:) that were on the 3TB HDD before as partions.
My UrBackup runs on a Linux (Unraid) machine in a Docker Container and the WebUI doesn’t provide me any options on those Backups. I tried to mount the vhd(z) files directly in the container (they seem to be stored as vhd but when I tried to mount them I got the following Error:
urbackupsrv mount-vhd --file Image_XXX_D.vhd --mountpoint /foo/D
Loading FUSE kernel module...
modprobe: FATAL: Module fuse not found in directory /lib/modules/5.10.28-Unraid
Starting VHD background process...
Waiting for background process to become available...
So this seems they are infact .vhdz and my container wasn’t compiled with fuse. It was built on arch afaik. So this seems to be not an option. What is the proper way to get all my files from B: and D: onto my external 5TB HDD? Please help me, I start to get really crazy because for me it looks like urBackup is only able to restore the the original HDD, I hope I am wrong?
Also, there is a thread discussing this from a few years back, it is mentioned you should be able to mount it using the webgui (I do not do img backups, so I can not test). Maybe you can do that if nothing else works so you at least has access to the data?
Thanks for the answer, I read that in the docs and for me it reads like it’s to combine multiple Volumes, so what I could do is combine Image of B: with Image of D: together with that, however that would exceed the 2TB limit of vhd. One vhd is like 1.78TB
When putting a single VHD, the docs state:
This tool can also be used to decompress images without decompressing their parents by selecting a single VHD file as input.
So if put the increment of a single volume, it will only use this and not decrompress all of it or am I getting this wrong or what does “not decompress parents” mean?
Thank you as well, sadly simply decompressing seems not to offer an option to give a destination path, it will do it in place. But the drive the backups are on has noth enough space left to have an extra 1.78TB one of my volumes has, so it appeared to me that isn’t an option, that’s why I am a little worried if I can get my files back
The web UI doesn’t offer me anything to mount, maybe because it’s running in a docker image, maybe because its setup like this, but for some reason I wasn’t able to find a way in the UI to do that, I can go to those saved volumes, but if I click on them I only get a small box of information, like name, size, etc. but no preview of files, no buttons or actions, I am out ofideas on that
Again, I do not do img backups, but I interpret it as: urbackupsrv -a /path/to/Image1.vhdz -a /path/to/Image2.vhdz\ -o /path/to/destination.vhd
I did not know about any size limitations on vhd files, sounds very strange. are you sure the limitation is not for fat32? (in cast there is a boot partition included or somtehing?) Or you were trying to write the file to a fat32 partition or something similar?
What error message do you get if you try the line I gave you? (edited to fit your case ofc)
Maybe I missunderstand completely how this works, do you have one or two .vhdz files for the B and E partitions?
Edit
You are correct, there is a 2TB limit on vhd files.
The VHD format has a built-in limitation of just under 2 TiB (2040 GiB) for the size of any dynamic or differencing VHDs. VHD (file format) - Wikipedia
Ha… have you tried just naming it vhdx?
It would be soooo stupid if that works…
It’s 2 files for volume B: and Volume D: one is 1.78TB and the other one like close to 1TB, both have a full image backup and some incrementals. Decompressing is not an option for each, because the decompress-file command only offers input file and not an output file, but will do it in place, which would double the size because each VHD would then exist twice, that rules this option out for me.
For assemble, that would also decompress if I put 2 volumes by the latest increment of each, it will create a new vhd file combining both, so it would create a vhd that is like 2.78TB which exceeds to limit of 2TB a VHD can have, so this rules that option out. And I don’t understand what happens on assembling a single Volume, it reads like it would decompress only the selected image, but not the parents, but espeically the parent is the big chunk I need…I don’t have an error for this, because I didn’t tried. I fear to much to crash the files, so I would only do it if I can be sure it will work.
I don’t think renaming the output to vhdx would work, it probably fails to create the vhd all together, so nothing to rename then. Also I doubt this works, I think vhdx are some special stuff, I didn’t see many virtual machine software allowing to create vhdx, most of them offer vhd or other formats, but with urbackup we are limited to vhd and vhdz
Yeah, the renaming was just a silly longshot idea. xD
According to the documentation, both methods would decompress
Consider using the method described in the next section 10.7 to decompress VHD files.
So: urbackupsrv -a /path/to/Image_B -o /path/to/output.vhd
for the first and then change to D (I do not know what your filenames are, but you probably understand)
I fail to see the risk of trying. If they are in separate files, try doing one at a time, mount that image in windows and then copy the files to the new hd. When done copying the files, unmount the file, remove it and then do the other partition and repeat copy.
I feel your data is not gone, you just have to figure out a way to get access to it by mounting them in windows.
As for how to mount in windows, maybe this can help?
There is also a link to a program in there called Kernel for VHD Recovery, but I would try to do the above first and if that doesn’t work, maybe look into a program like that?
You nailed exactly my current hurdle here, this is what I need to achieve, but first I somehow need to get a uncompressed VHD file that I can mount in windows.
Thank you for your help, maybe at some point I have to try it via assemble
I edited above after you responded so I repeat it here:
According to the documentation, both methods would decompress:
Consider using the method described in the next section 10.7 to decompress VHD files.
and at the end of that section:
This tool can also be used to decompress images without decompressing their parents by selecting a single VHD file as input.
So: urbackupsrv -a /path/to/Image_B.vhdz -o /path/to/output.vhd
for the first file, mount in windows and copy the files and then change to D (I do not know what your filenames are, but you probably understand) and repeat the copy.
Edit
It is my believe that images are a cool concept, but from reading the forum, imho it seems to be a big gamble to rely on. When you have a complete failure, you do NOT want to deal with this to restore.
If it is NOT something you HAVE to make images of (like you want to be able to get a bootable c drive back in order), why not just make file backups? That way you at least ALWAYS have access to the files directly, even if you are not able to use urbackup to restore them.
But that is just my opinion from reading about a lot of problems here on the forum when using images.
You are completely right, I wanted to try out urBackup and wasn’t sure how it works, all i wanted was to have a backup of my system. It’s like a year old, so it’s only about getting back what I can, if i’d knew everything will be locked inside of an vhd I would have done it differently, only now I learn how hard it seems to work with this. In the future I need to find better solutions for this, because I learned again how important backups are, because i stopped doing them because of storage limitations, I now need to live with the fact some data is gone, if i can restore properly from urback even everything is gone
There is a motto, or more a guideline in backup mentality that says: “a backup is not a backup until you have restored it” EXACTLY for this reason.
A backup is completely worthless, or even worse, it gives you a false sense of security unless you are ABSOLUTELY SURE you can restore from it, and the only way to find out is to actually restore. So that should be the first thing you do immediately after creating a backup. (not to the original location ofc but you have to make sure restoring is not an issue)
I feel pretty confident we can figure out a way to get your data back.
Try the above I suggested and come back with results. As long as you DO NOT DELETE THE FILES I’m pretty sure we can figure this out buddy!
I will do my best to try to help you!
Edit
The feeling of “I have lost it all” is TERRIBLE! I have been there and know where you are mentally, but try to stay calm. I REALLY get you and are here for you. <3
Thank you so much
For your kind words and your help, I really appreciate it <3
This is really some kind of a nightmare, I will try it and be back when I know more, but I have a question, in that post you wrote:
urbackupsrv -a /path/to/Image_B.vhdz -o /path/to/output.vhd
For some reason the file is not stored as vhdz but it’s still seems to be compressed, I just hope that doesn’t matter
Yeah, I was thinking that was a typo from you guys side, but since that is the case. Have you tried just mounting it (one of the non-incremental ones) directly in windows without decompressing it? Edit MAKE A COPY AND TRY TO MOUNT JUST IN CASE IT MESSES UP THE ORIGINAL FILE!!! You never know with windows.
It seems you have dropped the idea of trying to restore the image, witch I totally agree with. The goal now is to just get access to your data.
Edit
Is space limitation a problem here? do you have space on another drive where you can try these things out? Or maybe you can use the new 5TB and start with the smaller partition first to just see if it works.
I thought about trying something else, space is indeed an issue, but I hav an PC with a 4 TB HDD, I created a VM and I am adding 2 new fresh vhds to it for each Volume, so one at ~1TB and one at ~2TB (close to the limit) [That will take like 9 hours, 3 hours for the 1TB vhd and around 6 hours for the 1.99TB vhd] If that is done, I can start the vm with the restore CD on boot prio one and try to recover the backup into those vhds, but i fear it might take ages T_T
Why, just why
How is anyone supposed to take care of his data if it exceeds a few GB
I had a similar idea in my head if the mounting or decompressing did not work, but I was saving that for later since that is way more complicated than just mounting and copying the files…
Are you saying you want to make the virtual machine the simulated restored machine from the images?
I wouldn’t even make it that complicated, I would just spin up any virtual machine and then MOUNT the virtual hd:s (as extra hard-drives) in the machine and see if that gave me access to the data.
I recommend you try the simplest solutions first, THEN move on to more complicated stuff like that. As I said before, I think you should focus on just gaining access to the data rather than trying to restore the images.
BUT IN ANY CASE, MAKE SURE YOU DO NOT WORK AGAINST THE ORIGINAL FILES, MAKE COPIES AND WORK WITH THEM!!
Yes, that’s what I thought, because it mimics the usual way how to deal with a restore.
I get your point, but right now I am not sure what’s easier
The direct mounting wasn’t possible, as you can see in my opening post, that was the error when I tried to just mount them in the Docker Container UrBackup Server is running in, so my only option is to decompress them first somehow
Not working against the originals is difficult, they are so big if I copy them there is no room to decompress them on the same drive and it’s not like I have tons of such large HDDs laying around, sadly
Mount them on windows, you have established your installation was not built with fuse…
Yeah, that sux. I guess you just have to gamble it then if you cant fit one copy at a time on the 5TB drive and mount that in windows.
Besides, how do you access stuff on your linux from windows? Samba? Not sure how mounting a samba shared file on windows would work, that is why I keep insisting you copying the image to windows and trying to mount it there.
